ADAMTS20 (A disintegrin-like and metalloprotease domain with thrombospondin type-1 motifs) is a member of a family of secreted metalloproteases that can process a variety of extracellular matrix (ECM) components and secreted molecules. Adamts20 mutations in belted (bt) mice cause white spotting of the dorsal and ventral torso, indicative of defective neural crest (NC)-derived melanoblast development. The expression pattern of Adamts20 in dermal mesenchymal cells adjacent to migrating melanoblasts led us to initially propose that Adamts20 regulated melanoblast migration. However, using a Dct-LacZ transgene to track melanoblast development, we determined that melanoblasts were distributed normally in whole mount E12.5 bt/bt embryos, but were specifically reduced in the trunk of E13.5 bt/bt embryos due to a seven-fold higher rate of apoptosis. The melanoblast defect was exacerbated in newborn skin and embryos from bt/bt animals that were also haploinsufficient for Adamts9, a close homolog of Adamts20, indicating that these metalloproteases functionally overlap in melanoblast development. We identified two potential mechanisms by which Adamts20 may regulate melanoblast survival. First, skin explant cultures demonstrated that Adamts20 was required for melanoblasts to respond to soluble Kit ligand (sKitl). In support of this requirement, bt/bt;Kittm1Alf/+ and bt/bt;KitlSl/+ mice exhibited synergistically increased spotting. Second, ADAMTS20 cleaved the aggregating proteoglycan versican in vitro and was necessary for versican processing in vivo, raising the possibility that versican can participate in melanoblast development. These findings reveal previously unrecognized roles for Adamts proteases in cell survival and in mediating Kit signaling during melanoblast colonization of the skin. Our results have implications not only for understanding mechanisms of NC-derived melanoblast development but also provide insights on novel biological functions of secreted metalloproteases.
Mice with black and white coat coloration patterns have long been favorites of mouse fanciers and geneticists alike. Analysis of mouse coat color mutants has yielded important insights into normal developmental pathways as well as human disease processes. In this study we have investigated how mutations in a secreted metalloprotease, Adamts20, result in mice with white belts in their lumbar region, even though Adamts20 is not expressed in the pigment producing cells. Our findings suggest that the belting pattern is due to a combination of increased pigment cell death, decreased pigment cell number in the trunk, and functional overlap of closely related metalloproteases. Adamts20 mutants have disrupted function of Kit, a protein that regulates pigment cell development, as well as alterations in the extracellular matrix that surrounds the pigment cells. These findings have implications both for our understanding of general mechanisms of pigment cell development as well as for new biological functions of secreted metalloproteases.
Citation: Silver DL, Hou L, Somerville R, Young ME, Apte SS, Pavan WJ (2008) The Secreted Metalloprotease ADAMTS20 Is Required for Melanoblast Survival. PLoS Genet 4(2): e1000003. https://doi.org/10.1371/journal.pgen.1000003
Editor: Greg Barsh, Stanford University School of Medicine, United States of America
Received: September 21, 2007; Accepted: January 22, 2008; Published: February 29, 2008
This is an open-access article distributed under the terms of the Creative Commons Public Domain declaration which stipulates that, once placed in the public domain, this work may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose.
Funding: This work was supported by NIH award AR49930 (to S. Apte), NIGMS PRAT fellowship (to D. Silver), Intramural Research Program of NHGRI, NIH (to B. Pavan)
Competing interests: The authors have declared that no competing interests exist.
A disintegrin-like and metalloprotease with thrombospondin type-1 motifs (ADAMTS) metalloproteases constitute a large family of 19 zinc-dependent proteolytic enzymes that are distantly related to both the A disintegrin and metalloproteinase (ADAM) family, and to the matrix metalloproteinases (MMPs) ,. In contrast to the ADAM proteases that are membrane anchored, ADAMTS proteases are secreted; however, some may be considered as operational cell-surface proteases as they bind to the cell surface. Some ADAMTS proteases, such as ADAMTS10 (GeneID: 224697), ADAMTS13 (GeneID: 279028), and the procollagen amino-propeptidases (e.g. ADAMTS2), are highly specialized; others process a variety of substrates within the extracellular matrix (ECM), including chondroitin sulfate proteoglycans (CSPGs), such as aggrecan (GeneID: 11595) and versican (GeneID: 13003). Mouse models harboring mutations in ADAMTS family members have demonstrated the importance of these proteases during development. In some cases mutant phenotypes can be attributed to a failure to cleave specific substrates –. Mutation or dysregulation of ADAMTS proteases is associated with inherited and acquired pathologies including Ehlers-Danlos syndrome VIIC (OMIM#225410), thrombocytopenic purpura (OMIM#274150), Weill-Marchesani syndrome (OMIM#277600) and arthritis , –.
Among the animal mutants in Adamts proteases is a classical white-spotted mouse named belted (bt), so named because it contains white spots in the lumbar region creating the appearance of a belt –. Sequencing of three of the twelve known alleles of belted—btBei1, btMri1, and bt (Mouse Genome Informatics, MGI: 2660628)—revealed nonsense or missense mutations in Adamts20, thus implicating metalloproteases in skin pigmentation . Analyses of white spotting mutants suggest that such phenotypes are often due to defective development of neural crest (NC)-derived melanoblasts that produce pigment of the integument (skin, hair, feathers, and scales), inner ear, and eye ,.
Melanoblasts develop from a subset of NC that emigrate from the neural tube and overlying ectoderm and migrate dorso-laterally relative to the neural tube through prospective dermal mesenchyme (embryonic day (E) 8.5-E9.5) ,. Subsequently, the melanoblasts differentiate and expand (E9.5-E13.5), migrate into the epidermis and hair follicle (E13.5-E15.5), and eventually produce melanin (E15.5-P0). Several molecules, including the receptor tyrosine kinase Kit (GeneID: 16590) and its ligand, Kit ligand (Kitl, GeneID: 17311), regulate melanoblast development. Kit and Kitl act throughout melanoblast development, with independent requirements for melanoblast survival, proliferation and migration –. Kit is expressed on melanoblasts, and Kitl is expressed in the dermis and in dermal mesenchymal condensations and papillae , –.
Similar to Kitl, Adamts20 is expressed in dermal mesenchymal cells adjacent to and in advance of migrating melanoblasts throughout their development. This expression pattern led us to initially propose that the bt phenotype was caused by defective melanoblast migration . This hypothesis was also based upon the observation that the Adamts20 ortholog Gon-1 (GeneID: 177850) is essential for gonadal morphogenesis and distal tip cell migration in C.elegans –. The current study tests our hypothesis by performing extensive characterization of melanoblast development in bt/bt embryos. Our findings suggest that Adamts20 mutant mice exhibit white spotting in a belted pattern due to a combination of regional variation of melanoblast number, increased apoptosis, and functional overlap with Adamts9 (GeneID: 101401). Furthermore, our analyses of bt/bt embryos indicate that defective Kit signaling and versican processing may explain the failure of melanoblasts to develop properly.
Genetic and Molecular Characterization of a New Belted Allele
Adamts20 mutant mice exhibit white spotting in the lumbar region on both dorsal and ventral surfaces, frequently resulting in the appearance of a white belt in recessive animals (Figure 1A) –. For these studies an allele of bt on an inbred C57BL/6 background was used, as different genetic backgrounds can affect the expressivity of the bt phenotype ,. Since an Adamts20 mutation had not been specifically demonstrated in this particular bt allele, a complementation cross was performed between these bt mice and btBei1/+ mice. Seven out of 10 animals born exhibited a bt phenotype, consistent with the expected Mendelian ratios (50% bt) and indicative of a recessive allele of bt. This allele has now been designated bt9J and is listed at MGI.
(A) An image of a bt9J/bt9J mouse (left) and a C57BL/6 mouse (right). (B) A representative result from a genotyping assay for the following genotypes: bt9J/bt9J (red triangle), bt9J/+ (green square), wild-type (C57BL/6) (blue diamond), and water control (grey circle).
The Adamts20 gene was sequenced from bt9J/bt9J genomic DNA in order to identify the molecular lesion. A single C to T point mutation was identified at nucleotide 2451, which is predicted to cause a missense mutation (Leu761Phe) in the spacer domain of ADAMTS20. The leucine residue mutated in bt9J mice is present in both mouse and human Adamts20 genes and is highly conserved among all Adamts mouse genes (present in 15 of the 19 genes). This suggests that Leu761 may be a critical residue for ADAMTS20 folding or for enzymatic function. We designed a TaqmanTM assay to genotype bt9J mice (see Materials and Methods). We determined that bt9J is the same strain as Mutant Mouse Resource Center (MMRC) strain #183 (Figure 1B). Although the missense mutations in bt9J, btMri1, and bt affect different domains of ADAMTS20, these and other bt alleles are recessive and exhibit very similar phenotypes, strongly suggesting they act as functional null alleles.
Initial Specification and Migration of Melanoblasts Is not Disrupted in bt9J/bt9J Embryos
White spotting of mouse coats is typically caused by defective melanoblast development during embryogenesis. In order to elucidate when and how Adamts20 affects melanoblast development, melanoblast distribution in bt9J/bt9J embryos was examined. We generated C57BL/6 mice containing a Dct-LacZ transgene, which marks specified melanoblasts, and crossed these onto a C57BL/6 bt9J/bt9J background ,. Since there are no phenotypes associated with bt9J/+ adult animals, this heterozygous genotype served as a control for this study.
We compared melanoblast distribution in whole mount bt9J/+;Dct-LacZ and bt9J/bt9J;Dct-LacZ embryos at E11.5 through E16.5 (Figure S1 and Figure 2). In E11.5 and E12.5 control embryos, melanoblasts had completed their initial migration from the neural tube and were distributed evenly across the dorsal and lateral surfaces of the embryo (Figures S1 and Figure 2A and 2C). A similar distribution pattern in E11.5 and E12.5 bt9J/bt9J embryos was observed (Figures S1 and Figure 2B and 2D). Quantification of melanoblasts in the head and the presumptive belt region of the trunk of E12.5 whole mount embryos showed no difference in melanoblast number between control and bt9J/bt9J embryos (Table 1, boxes in Figure 2A and 2C). These results show that initial specification and migration of NC-derived melanoblasts is not disrupted in bt9J/bt9J animals.
(A–P) Images of β-galactosidase stained whole mount bt9J/+;Dct-LacZ (A,C,E,G,I,K,M,O) and bt9J/bt9J;Dct-LacZ embryos (B,D,F,H,J,L,N,P). Shown are E12.5 embryos (A,B), E12.5 trunks (C,D), E13.5 embryos (E,F), E13.5 trunks (G,H), E14.5 embryos (I, J), E14.5 trunks (K, L), E15.5 trunks (M,N) and E16.5 embryos (O,P). Each image is representative of at least 12 embryos. Boxes in A, C, E, G, I, K indicate regions in which melanoblasts were quantitated; the indicated regions were quantitated in both bt9J/+;Dct-LacZ and bt9J/bt9J;Dct-LacZ embryos (see Table 1). Scale bars are: (A,B,E,F,I,J,O,P) 2 mm, (C,D,G,H,K,L,M,N) 1 mm.
Melanoblast Distribution Is Reduced only in the Trunk of bt9J/bt9J Embryos
Since there was no defect in younger embryos, melanoblast distribution in bt9J/+; Dct-LacZ and bt9J/bt9J;Dct-LacZ E13.5-E16.5 embryos was compared (Figure 2E–2P). In E13.5 and E14.5 bt9J/+;Dct-LacZ embryos, the melanoblast population had expanded throughout the embryo and melanoblasts were distributed uniformly across dorsal and lateral surfaces of the trunk (Figure 2E, 2G, 2I, and 2K). In contrast, melanoblast distribution was specifically reduced in the trunk of E13.5 and E14.5 bt9J/bt9J;Dct-LacZ embryos (Figure 2F, 2H, 2J, and 2L). Quantitation of the head and trunk region of E13.5 and E14.5 bt9J/bt9J;Dct-LacZ embryos confirmed that the melanoblast defect was limited to the trunk (Table 1, boxes in Figure 2E, 2G, 2I, 2K) (p = 0.0000002 at E13.5, p = 0.000096 at E14.5). By E15.5 and E16.5, melanoblasts were distributed evenly on both dorsal and ventral surfaces of bt9J/+;Dct-LacZ embryos (Figure 2M and 2O). However bt9J/bt9J;Dct-LacZ embryos displayed a notable absence of melanoblasts on both surfaces in the region corresponding to the future belt (Figure 2N and 2P). Melanoblast distribution in the head and tail regions was similar between E16.5 bt9J/+;Dct-LacZ and bt9J/bt9J;Dct-LacZ embryos (Figure 2O and 2P). Close examination of the belt region in E16.5 bt9J/bt9J;Dct-LacZ embryos showed no accumulation of melanoblasts at the edges of the belt, as might be expected if migration into this region were impaired (Figure S2). Collectively, these analyses of the melanoblast distribution in bt9J/bt9J animals reveal the following: the phenotype is first apparent at E13.5, melanoblasts are reduced only in the trunk region (the location of the presumptive belt), and dorso-lateral melanoblast migration is not impaired.
Melanoblasts Are Present in All Skin Compartments of bt9J/bt9J Embryos
The melanoblast defect is first observed at E13.5, coincident with the timing of melanoblast migration from the dermis into the epidermis, and with generalized dermal expression of Adamts20 ,. Therefore we hypothesized that Adamts20 may be required for normal distribution of melanoblasts in dermal and epidermal compartments of the skin. To address this, melanoblasts were quantified in the dermis, epidermis, and dermal-epidermal border in E13.5 trunk sections (Figure 3A and Table 2) (see Materials and Methods). In control trunk sections (n = 178) we observed an average of 33.6 melanoblasts per section, whereas in bt9J/bt9J;Dct-LacZ sections (n = 179) there was an average two-fold reduction in melanoblast number (15.9 per section, p<0.0001) (Table 2). This finding is consistent with the analyses of whole mount embryos (see Table 1 and Figure 2E–2H). However, the relative distribution of melanoblasts within each of the skin layers in bt9J/bt9J;Dct-LacZ embryos did not differ significantly from bt9J/+;Dct-LacZ control embryos (p = 0.056) (Figure 3B). These results show that while melanoblast number is significantly reduced in the lumbar region of bt9J/bt9J embryos, melanoblast migration from the dermis into the epidermis is unaffected.
(A) A cartoon representation of an E13.5 whole mount embryo showing the region between the fore- and hind-limbs from which 16 µm trunk sections were analyzed. (B) Pie charts representing the melanoblast distribution in the dermis (blue), epidermis (red), and dermal/epidermal border (yellow) from E13.5 bt9J/+ (top) and bt9J/bt9J (bottom) sections with the total number of melanoblasts indicated in parentheses (n = 2 embryos, each genotype). There was no significant difference in melanoblast distribution between bt9J/+ and bt9J/bt9J embryos (p = 0.056).
bt9J/bt9J Embryos Exhibit Normal Melanoblast Proliferation but Defective Melanoblast Survival
At E13.5, when the defect in melanoblast development in bt9J/bt9J embryos is first apparent, melanoblasts are undergoing dramatic increases in cell proliferation . To assess if the bt phenotype is caused by reduced melanoblast proliferation in the trunk region of bt9J/bt9J embryos, the mitotic index of melanoblasts at E13.5 was examined in bt9J/+;Dct-LacZ and bt9J/bt9J;Dct-LacZ trunk sections (Figures 3A and 4). Melanoblasts undergoing mitosis were identified by co-expression of β-galactosidase and phospho-histoneH3 (PH3) (Figure 4A–4C). The average percentage of dividing melanoblasts in control and bt9J/bt9J sections was not significantly different (2.94% and 3.03% respectively, p = 0.9067) (Figure 4G), suggesting that the melanoblast reduction in bt9J/bt9J embryos is not a consequence of abnormal proliferation.
The reduced number of melanoblasts in the trunk region of E13.5 bt9J/bt9J;Dct-LacZ embryos could also be due to a failure of melanoblasts to survive . To examine this, the apoptotic index of melanoblasts was quantitated from cells co-expressing β-galactosidase and cleaved-caspase3 (CC3) (Figures 3A and 4D–4F). The average percentage of apoptotic melanoblasts was significantly increased in bt9J/bt9J;Dct-LacZ sections relative to bt9J/+;Dct-LacZ sections (compare 8.4% to 1.2%, p<0.00005) (Figure 4H, Table 2). Apoptotic melanoblasts were apparent in both dorsal and ventral skin in control and mutant embryos. In addition, no significant differences were seen in the distribution of apoptotic, CC3+melanoblasts in the three skin compartments, as follows: dermal (bt9J/bt9J = 77.8%; control = 82.4%), epidermal (bt9J/bt9J = 14.8%; control = 5.8%), and dermal/epidermal (bt9J/bt9J = 7.4%; control = 11.85%), (p = 0.6133). These results show that Adamts20 is required for melanoblast survival in all cell layers of the trunk at E13.5.
Adamts20 is expressed along the length of the embryo, yet the spotting phenotype in adult mice is evident only in the lumbar region. Therefore we examined melanoblast number and melanoblast apoptosis in a region outside of where the belt occurs, in the head (Table 2). In normal embryos, the density of melanoblasts is not the same in all regions, as previously described –. As expected, quantitation of melanoblasts in control embryos demonstrated that there were over three-fold more total melanoblasts in the head region than in the trunk (compare 110 versus 33.6, p = 0.0001). There was no difference in melanoblast number in the head of control and bt9J/bt9J;Dct-LacZ embryos (Table 2), similar to what was seen in whole mount analyses.
The apoptotic index of melanoblasts varied in different regions of control embryos as well as between control and mutant embryos (Table 2). In control embryos, melanoblast apoptosis was lower in the head than in the trunk. Comparing control and bt9J/bt9J embryos, we found that apoptosis was increased in the head although to a lesser extent than observed in the trunk. This result indicates that Adamts20 is required for melanoblast survival throughout the embryo even though it does not result in white spotting in the head.
Melanoblasts in bt9J/bt9J Skin Fail To Respond to sKitl
Melanoblast survival is dependent upon Kit activation , by Kit ligand (Kitl), which is expressed in the dermis in a similar temporal and spatial pattern to Adamts20 , , –. We hypothesized that Adamts20 regulates melanoblast survival through modulation of Kit signaling, and assessed the effects of altered Kit signaling upon the extent of white spotting in bt9J/bt9J animals harboring various mutant alleles of Kit (MGI: 96677) or Kitl (MGI: 96974). Kittm1Alf/+ mice contain a null mutation in Kit and have primarily ventral spotting –. As depicted in Figure 5, bt9J/bt9J;Kittm1Alf/+ mice exhibited dramatically wider dorsal and ventral belts in comparison to bt9J/bt9J mice or Kittm1Alf/+ mice alone (Figure 5A and 5B). The increased spotting was a synergistic effect rather than an additive one (compare 46.9% of dorsal and ventral surfaces in bt9J/bt9J;Kittm1Alf/+ with 11.2% and 9.1% in Kittm1Alf/+ and bt9J/bt9J, respectively). Synergistic increases in spotting were also observed in bt9J/bt9J mice carrying mutant alleles of Kitl. Heterozygosity for either KitlSl, a null allele, or KitlSl-d, a deletion that generates short soluble Kitl but not membrane–bound Kitl, combined with homozygosity for bt9J resulted in synergistically increased spotting (Figure 5C and 5D) ,. In contrast, bt9J/bt9J mice carrying a mutation in Mitf (bt9J/bt9J;MitfMi/+) (MGI: 104554) exhibited no synergistic spotting (Figure 5E). These results show that decreasing Kit signaling exacerbates the bt9J/bt9J phenotype.
(A–F) Representative immunofluorescence images depicting 16 µm trunk sections from bt9J/bt9J E13.5 embryos assayed for proliferation (A–C) and apoptosis (D–F). Proliferating melanoblasts were identified by staining sections for LacZ (cytoplasmic red, A) and PH3 (nuclear green, B), and their co-localization together with nuclear DAPI staining (blue, C). Apoptotic melanoblasts were identified by staining sections for LacZ (cytoplasmic red, D) and CC3 (cytoplasmic green, E), and their co-localization together with DAPI (blue, F). (G–H) The Y axis (%Dct+PH3+ and %Dct+CC3) indicates the average percentage of total melanoblasts per section that are PH3 positive (G) or CC3 positive (H), for bt9J/+ (blue) and bt9J/bt9J (red) E13.5 embryos (n = 5 embryos for each genotype). The total number of sections and melanoblasts analyzed are listed beneath each graph, as well as the standard deviations. Asterisks indicate highly significant P values. The scale bars are: (A–F) 0.5 µM.
(A–E) Shown are representative dorsal and ventral images of mice (top and middle row, respectively). The genotypes and number of animals (n) analyzed were: (A) bt9J/bt9J (n = 31), (B) Kittm1Alf/+ (n = 14) and bt9J/bt9J;Kittm1Alf/+ (n = 15), (C) KitlSl/+ (n = 24) and bt9J/bt9J;KitlSl/+ (n = 17), (D) KitlSl-d/+ (n = 6) and bt9J/bt9J;KitlSl-d/+(n = 6), and (E) MitfMi/+ (n = 8), and bt9J/bt9J;MitfMi/+ (n = 13). Below are graphs showing quantitation of the average proportion of white spotting with standard deviations. P values represent comparisons of spotting between compound mutants and bt9J/bt9J animals, and between compound mutants and Kit mutants. Asterisks indicate highly significant values.
Heterozygous KitW-v mutations reduce melanoblast number beginning at E10.5 , which is three days prior to the onset of the melanoblast defect in bt9J/bt9J. To determine if decreasing Kit signaling in bt9J/bt9J embryos would lead to an earlier defect in melanoblast development than seen in bt9J/bt9J alone, we examined E12.5 bt9J/+;Kittm1Alf/+ and bt9J/bt9J;Kittm1Alf/+ embryos (Figure S3). Since the Kittm1Alf mutation is a LacZ knock-in at the Kit locus, we monitored melanoblasts using β-galactosidase staining. There was no detectable difference in Kit-positive melanoblast distribution between bt9J/+;Kittm1Alf/+ and bt9J/bt9J;Kittm1Alf/+ embryos at E12.5. These results demonstrate that the synergistic defect in melanoblast development does not precede the onset of the bt phenotype at E13.5.
The synergistic interaction studies indicate that Kit signaling may be disrupted in Adamts20 mutant animals. To test if melanoblasts in bt9J/bt9J embryos could respond to sKitl we used an ex vivo embryonic skin explant assay using dorsal trunk skin from E13.5 embryos –. Observation of bt9J/+;Dct-LacZ and bt9J/bt9J;Dct-LacZ skin after four days in culture demonstrated that ex vivo explant culture conditions could recapitulate both normal melanoblast colonization and the bt defect (Figure 6). Melanoblasts in the control skin entered hair follicles and were distributed evenly across the explant (n = 5) (Figure 6A). While some melanoblasts survived and migrated into hair follicles in the bt9J/bt9J;Dct-LacZ skin, in a large domain of the explant the melanoblasts were reduced (n = 5) (Figure 6B), similar to the phenotype of whole mount E16.5 bt9J/bt9J;Dct-LacZ embryos (Figure 2O and 2P).
(A–D) Images of β-galactosidase stained skin cultures grown from E13.5 bt9J/+;Dct-LacZ (A,C) or bt9J/bt9J;Dct-LacZ (B,D) embryos, and treated with either 5% FBS (A,B) or 5% FBS+Kitl (C,D). Brackets indicate the presumptive belt regions in bt9J/bt9J skin. The scale bars are: (A–D) 0.5 mm.
Previous studies showed that melanoblasts within embryonic skin cultures respond to soluble Kitl (sKitl)  and that sKitl promotes melanoblast survival in NC cultures . ADAMTS20 could be required for proteolytic cleavage of Kit or Kitl, or for modifying additional molecules necessary for Kit signaling. To test if Kit signaling was defective in bt9J/bt9J embryos, we asked whether melanoblasts in bt9J/bt9J embryos could respond to sKitl and whether sKitl could rescue the bt phenotype. Similar to studies by Jordan et. al, an increase in melanoblast number and colonization of hair follicles was observed in control skin exposed to sKitl as compared to FBS alone (Figure 6C, n = 5). In contrast, exposure of bt9J/bt9J;Dct-LacZ skin (n = 7) to sKitl did not restore melanoblasts to the presumptive belt (Figure 6D). Interestingly, melanoblasts at the rostral and caudal regions of the trunk did not increase in number, even though these are outside of the expected belt. These results show that in the absence of Adamts20, melanoblasts are unable to respond to sKitl, indicating that Kit signaling is disrupted in bt9J/bt9J mutants. Furthermore, since sKitl could not rescue the bt phenotype, it suggests that the melanoblast defect in bt9J/bt9J animals is not caused by defective cleavage of Kitl.
ADAMTS20 Cleaves Versican in vitro and Is Necessary for Versican Cleavage in vivo
We explored the possibility that Adamts20 could regulate melanoblast survival through additional pathways. One of the best-understood activities of ADAMTS proteases is processing of various ECM substrates. Of note, ADAMTS20 is most closely related phylogenetically to ADAMTS metalloproteases that specifically process chondroitin sulfate proteoglycans (CSPGs) ,,. Although no substrates for ADAMTS20 have been identified, the CSPG versican is enriched in the skin and implicated in NC development, making it an excellent candidate substrate of ADAMTS20 –.
We examined Versican and Adamts20 expression by whole mount in situ hybridization in wild-type embryos. At E12.5, when melanoblasts are widely distributed in the trunk, both Versican and Adamts20 are expressed broadly across the trunk, overlapping in many regions of the skin and the developing mammary ducts (data not shown). Since both versican and ADAMTS20 are secreted proteins, the expression patterns of their RNAs relative to each other and to melanoblasts support the possibility of a functional relationship in vivo.
We assessed the ability of ADAMTS20 to process versican by expressing ADAMTS20 in 293 cells, and subsequently incubating these cells with versican (Figure 7A). Versican is alternatively spliced to generate four isoforms, with VersicanV1 being the dominant variant in adult skin –. Versican cleavage was examined using anti-DPEAEE antibody, which specifically recognizes the neo-epitope generated by ADAMTS cleavage of versicanV1 (70 KDa) . While secreted media from 293 cells alone exhibited little to no cleavage of versican, versican processing was evident in secreted media from 293 cells expressing Adamts20 (Figure 7A). As a positive control, cleaved versican was also evident in secreted media from cells expressing Adamts9, which has previously been shown to process versican . These results show that in vitro, versican is a substrate for cleavage by ADAMTS20.
(A,B) Western analyses probed with anti-DPEAEE that recognizes the 70 KDa versican cleavage product. (A) Protein extracts from 293 cells transfected with vector alone, Adamts20 (2 experiments) and Adamts9, and incubated with versican. (B) Protein extracts of E15.5 day old skin from six different bt9J/+ and bt9J/bt9J embryos. Anti-α-Tubulin serves as a loading control and anti-GAG βindicates levels of total versican. (C–J) Immunofluorescence images of +/+ (C,E,G, I) or bt9J/bt9J (D,F,H,J) E15.5 10 µm trunk (C–H) and forelimb (I,J) sections stained with anti-DPEAAE (C,D,I,J), anti-DPEAAE and DPEAEE peptide (E,F), and anti-GAG β (G,H). The anti-DPEAEE staining in the proliferating layer of the epidermis (indicated by arrows) is reduced dramatically in trunk sections and reduced to a lesser extent in forelimb sections from bt9J/bt9J embryos. The scale bars are: (C–F,I,J) 50 µM, (G,H) 100 µM.
Analysis of skin extracts from the dorsal trunk of E15.5 bt9J/+ and bt9J/bt9J embryos showed ADAMTS20 is also necessary for versican cleavage in vivo (Figure 7B). The levels of the cleaved 70 KDa band were reduced in bt9J/bt9J extracts compared to control extracts (n = 4 experiments). We also observed a reduction in levels of a 220 KDa band corresponding to the V0 isoform (data not shown). In contrast, there was no alteration in total versican, which was assessed using a GAG β antibody that recognizes the GAG β domains of full-length versicanV0 and V1 (Figure 7B). Collectively, these results show that ADAMTS20 cleaves versican in vitro and that it is required for versican cleavage in skin in vivo.
Immunofluorescence was performed to evaluate the spatial pattern of versican cleavage in embryonic skin. Versican expression was assessed on trunk sections of E15.5 C57BL/6 (+/+) and bt9J/bt9J embryos (Figure 7C–7H). In +/+ sections, cleaved versican was evident in the dermal mesenchyme and was enriched in the proliferating basal layer of the epidermis (arrow in Figure 7C) but was excluded from the outermost layer of the epidermis, in a similar pattern to that of intact versican in adult human skin . In contrast to +/+ sections, bt9J/bt9J sections displayed a dramatic reduction in cleaved versican in the epidermis (Figure 7D). Inclusion of the DPEAEE peptide in the staining procedure demonstrated the specificity of the antibody for this peptide (Figure 7E and 7F). Total versican expression was similar in both +/+ sections and bt9J/bt9J sections, with high expression in the basement membrane, dermis, and dermal condensations, but lower levels in the epidermis (Figure 7G and 7H). Outside of the belt region, at the level of the forelimb, the levels of cleaved versican in bt9J/bt9J sections were reduced relative to wild-type, but not to the same extent as seen in the trunk (Figure 7I and 7J). Together these results show that in bt9J/bt9J embryos, levels of processed versican, but not total versican, are reduced in the skin of the embryonic trunk, and show that ADAMTS20 is necessary for versican remodeling during melanoblast development.
Adamts9 and Adamts20 Act Cooperatively To Regulate Melanoblast Development
Although Adamts20 is expressed across the length of the embryo , only the lumbar region of the trunk in bt9J/bt9J mutants shows dramatic reduction of melanoblasts and white spotting. Given that ADAMTS20 belongs to a large family of closely related metalloproteases, we reasoned that partial redundancy with other metalloproteases might account for the higher proportion of melanoblasts surviving outside of the lumbar region. The two closest Adamts20 homologs, Adamts5 (GeneID: 23794) and Adamts9, are both expressed in the embryonic skin along the length of the embryo and are both versicanases, making them excellent candidates for regulating melanoblast development ,,(Bon-Hun Koo and S.S. Apte, unpublished data). To test if these metalloprotease genes were functionally redundant with Adamts20, we evaluated the extent of white spotting in mutants lacking Adamts20 and either Adamts9 (Adamts9ko) or Adamts5 (Adamts5ko).
Adamts9ko homozygotes are lethal prior to gastrulation (H. Enomoto and S. Apte, submitted) therefore we used Adamts9ko heterozygotes to examine a functional overlap between Adamts9 and 20. While mice heterozygous for mutation of Adamts20 and 9 are viable, bt9J/bt9J;Adamts9ko/+ animals die shortly after birth (H. Enomoto and S. Apte, submitted). We examined pigmentation in newborn animals that were wild-type, bt9J/+;Adamts9ko/+, bt9J/bt9J, or bt9J/bt9J;Adamts9ko/+. Skin pelts containing dorsal and ventral surfaces and spanning the entire trunk region were removed from newborn pups (Figure 8A). Although not externally obvious until 4–5 days after birth, melanoblasts can be seen as clusters of pigmented cells on the dermal side of the P0 epidermis. In all skin samples, pigmentation was apparent primarily on dorsal surfaces as shown by schematic representations in Figure 8B–8E. In the skin from +/+ and bt9J/+;Adamts9ko/+ animals, pigmentation was distributed uniformly between anterior and posterior boundaries of the skin samples (n = 4 each) (Figures 8B and 8C). In the bt9J/bt9J skin samples (n = 6), a partial de-pigmentation was observed on about 50% of the anterior-posterior length, consistent with the likely boundaries of “belted” and “non-belted” regions (Figure 8D). However, in the bt9J/bt9J;Adamts9ko/+ skin samples (n = 5), de-pigmentation was apparent on 75% of the anterior-posterior length (Figure 8E). Our analyses showed a statistically significant increase in the severity of de-pigmentation in bt9J/bt9J;Adamts9ko/+ skin compared to that seen in bt9J/bt9J skin (Figure 8F). In contrast, the pigmentation phenotype in btBei1/btBei1;Adamts5ko/Adamts5ko mice was similar to that of btBei1/btBei1 animals indicating that loss of Adamts5 does not exacerbate the bt9J/bt9J phenotype and that unlike Adamts9, Adamts5 does not participate in melanoblast development (D. McCullough and S. Apte, unpublished data).
(A) A schematic image of a newborn pup. Skin samples (indicated by the box) were taken from the trunk region around the circumference of the animal and were flattened with a single cut on the ventral side. (B–E) Shown on the left are cartoon representations of the pigmentation patterns and on the right are high magnification images of skin samples from (B) +/+ (n = 4), (C) bt9J/+; Adamts9ko/+(n = 4), (D) bt9J/bt9J (n = 6), and (E) bt9J/bt9J; Adamts9ko/+(n = 5) animals. Dots indicate pigmented hair follicles. Labels for orientation of the samples are as follows: V (ventral), D (dorsal), A (anterior), and P (posterior). Images were taken by focusing on 1/3 of the anterior-posterior length of the skin sample. (F) A graph depicting the relative score of pigmentation as described in the Materials and Methods. The difference between bt9J/bt9J and bt9J/bt9J; Adamts9ko/+ samples was significantly different (p = 0.006). (G and H) Head images of in situ hybridization with the melanoblast marker, Pmel17, on whole mount E13.5 bt9J/+; Adamts9ko/+(n = 2) (G) and bt9J/bt9J; Adamts9ko/+(n = 2) (H) embryos. Note there are fewer melanoblasts all over the head and particularly in the snout of bt9J/bt9J; Adamts9ko/+embryos. Scale bars = 0.5 mm.
In order to assess melanoblast development in bt9J/+;Adamts9ko/+and bt9J/bt9J;Adamts9ko/+ mice, E13.5 embryos were examined using in situ hybridization with Pmel17 (GeneID: 20431), a marker specific for melanoblasts. Similar to what was observed in newborn skin, there was a significant reduction in melanoblasts in the trunk region of bt9J/bt9J;Adamts9ko/+ embryos compared to bt9J/+;Adamts9ko/+ embryos (data not shown). Importantly, there was also a significant reduction of melanoblasts in regions outside of the trunk (Figure 8G and 8H). These results indicate that Adamts9 and Adamts20 are partially redundant during melanoblast development not only in the trunk but in the head as well.
Analysis of mouse coat color mutants has yielded mechanistic insights into normal developmental pathways as well as human disease processes. In this paper we explored how mutation of Adamts20, a secreted metalloprotease that is not expressed in melanoblasts, can disrupt melanoblast development and cause white spotting. The expression pattern of Adamts20, along with studies performed on the C.elegans ortholog Gon-1, initially suggested that Adamts20 regulated melanoblast migration , –. However, rather than this predicted requirement for melanoblast migration, we have identified an unexpected role for Adamts20 in melanoblast survival. Our results have implications not only for understanding mechanisms of melanoblast development and pigmentation-associated diseases but also for the biological functions of secreted metalloproteases.
Mutation of Adamts20 Increases Melanoblast Apoptosis
The most striking alteration observed in bt9J/bt9J embryos was a reduction in trunk melanoblasts beginning at E13.5. Early events of melanoblast specification, migration, proliferation and survival were not affected, as the number and distribution of Dct-LacZ-positive melanoblasts was similar between control and bt9J/bt9J embryos up to E12.5. This result contrasts many melanoblast mutants including Kit, Mitf, Pax3 (MGI: 97487), Ednrb (MGI: 102720), and Sox10 (MGI: 98358), in which melanoblast defects can be observed as early as E10.5-E11.5 .
Multiple lines of evidence also indicate that later stages of melanoblast migration were not defective in bt9J/bt9J embryos. Dct-LacZ-positive cells did not accumulate dorsally or in the dermis, suggesting that melanoblast migration along the dorso-lateral pathway and between skin layers was normal. It is also unlikely that melanoblasts undergo cell death as a result of or subsequent to failed migration, as apoptotic melanoblasts were apparent in both dorsal and ventral regions of mutant embryos and in similar proportions within dermal and epidermal compartments. Our results also exclude a defect in the second wave of melanoblast migration ,, since melanoblasts did not accumulate at the belt edges in E16.5 embryos. Instead, our results show that reduction of melanoblasts is due to the seven-fold increase in apoptosis observed in the trunk region of Adamts20 mutant embryos.
Why Do Mutations in Adamts20 Cause a Belt?
We propose that the white belt occurs in Adamts20 mutants for a combination of three factors: uneven densities of melanoblasts along the length of normal embryos, regional differences in apoptosis, and functional redundancy with other metalloproteases.
In wild-type animals there are fewer embryonic melanoblasts in the trunk relative to other areas of the body, as shown in this study and by others , –37. Related to this, many of the coat color mutants exhibit spotting in the lumbar level of the trunk suggesting that melanoblasts in this region are particularly sensitive to genetic perturbation, perhaps due to reduced numbers. This suggests that if melanoblast numbers were further reduced along the trunk in Adamts20 mutants, the resulting belt would be wider. Indeed this was the case since genetic interactions of bt9J/bt9J with Kit mutations caused a wider belt as opposed to increased spotting in the head.
Second, Adamts20 is required for melanoblast survival throughout the embryo, consistent with the expression pattern of Adamts20 along the length of the embryo. In bt9J/bt9J embryos, we found significant increases in apoptosis in both head and trunk regions. Importantly, the proportion of melanoblasts undergoing apoptosis was much more pronounced in the trunk than in the head. Thus a combination of increased apoptosis and lower initial melanoblast number in the trunk region of a bt9J/bt9J mutant depletes melanoblast numbers to levels at which it cannot generate adequate pigmentation.
Third, Adamts20 mutant animals have limited white spotting due to functional redundancy with other metalloproteases, including Adamts9. In a bt9J/bt9J background, loss of a single copy of Adamts9 results in reduced melanoblast numbers in the trunk as well as in the head. Thus, in bt9J/bt9J mice, Adamts9 can compensate for Adamts20 deficiency in regions of higher melanoblast number such as the head. The early embryonic lethality of Adamts9ko/Adamts9ko mice makes it impossible to currently assess the absolute requirement of Adamts9 in pigmentation. However, the widespread expression of Adamts9 in the skin paired with the dramatic defect in melanoblast development in bt9J/bt9J;Adamts9ko/+ animals predict an even greater pigmentation defect in a homozygous background. These future analyses will depend upon the generation of a conditional Adamts9 knock-out in the skin.
Adamts20 Modulates Kit Signaling
Our results suggest that Adamts20 is required for melanoblast survival at least in part through modulating Kit signaling. We show that bt9J/bt9J;Kittm1Alf/+ mutants exhibit synergistic spotting, and melanoblasts in bt9J/bt9J skin cultures are unable to respond to sKitl. This mechanism is consistent with the requirement of Kit signaling for melanoblast survival as well as the overlapping expression of Adamts20, Kit and Kitl in the dermis , , , –,.
Kit signaling is essential for several events in melanoblast development beginning at E10.5 and including both migration and survival. Yet our findings indicate that mutation of Adamts20 modulates only a subset of Kit functions, namely survival, and only during a defined window of time, around E13.5, rather than throughout development. This is evidenced by the fact that bt9J/bt9J embryos do not show altered melanoblast development prior to E13.5, such as seen in Kit/+ embryos ,. In addition, Kit heterozygous mutations do not exacerbate the onset of the melanoblast defect in E12.5 bt9J/bt9J embryos, as would be expected if Adamts20 were required earlier. The stage at which Adamts20 disrupts melanoblast development may also explain why bt mice do not exhibit extensive ventral spotting like other white spotted mutants that impede earlier events in development.
Given these observations, it is notable that restricted disruption of Kit, during the embryonic period associated with onset of the bt9J/bt9J phenotype, results in mice with spotting phenotypes similar to bt ,. Pregnant mice injected with neutralizing Kit antibodies between E10.5 and E13.0 give birth to mice exhibiting white belts. In contrast, injection of these Kit antibodies into pregnant animals prior to E10.5 and after E13.0 results in mostly unpigmented mice.
A white belt/band is also evident in KitW-sh/+, KitW-57J/KitW-57J, and KitW-bd/+ mice that carry mutations in Kit regulatory sequences and cause altered Kit expression patterns ,,. In addition, Kimura and colleagues describe belted phenotypes in Kittm1Ber and Kittm2Ber mice containing targeted mutations of Kit at tyrosine residues 567 and 569, amino acids essential for proper phosphorylation and downstream signaling . It is interesting that disruption of these specific amino acids results in spotting similar to that seen in belted mice, and we speculate that Adamts20 may be required for efficient Kit phosphorylation at these sites in melanoblasts in vivo.
There are several possible mechanisms by which ADAMTS20 could regulate Kit signaling. ADAMTS20 could be required directly for cleavage of either Kit and/or Kitl to produce sKitl. For example, the ADAM family member ADAM17 (GeneID: 11491) cleaves Kit in vitro, and ADAM17, 19 (GeneID: 8728) and 33 (GeneID: 110751) cleave Kitl –. However, sKitl did not rescue the melanoblast defect in bt9J/bt9J skin cultures as would be expected if Kitl were an essential substrate of ADAMTS20. As further evidence that the bt phenotype is not due to defective Kitl cleavage, bt9J/bt9J exhibited similar genetic interactions with both a Kitl null allele and with KitLSl-d/+ animals, which produce a short sKitl ,,. Kunisada et. al have also shown that transgenic mice over-expressing Kitl transgene, which greatly reduces the white spotting associated with Kit mutants, only slightly reduces the spotting in bt9J/bt9J animals . Taken together this suggests that the bt phenotype is not due to defective cleavage of Kitl. Instead, Adamts20 may regulate the interaction of Kitl with Kit, the activation of Kit, or signaling downstream of the receptor. Future biochemical studies will be necessary to define the exact mechanism by which Adamts20 modulates Kit signaling.
The observation that melanoblasts in bt9J/bt9J trunk explants did not exhibit any response to soluble Kitl indicates that Kit signaling is defective, but it is important to note that this does not exclude the possibility that Adamts20 may additionally regulate other factors essential for melanoblast development. For example, Adamts20 could be required to activate other pathways that act synergistically with Kit to regulate melanoblast survival.
ECM Alterations in bt9J/bt9J Mice
Another possible mechanism by which Adamts20 could modulate survival is by altering the extracellular matrix in which melanoblasts are located. We show that bt9J/bt9J mutants exhibit reduced cleavage of at least one ECM component, versican. Versican, Adamts20, and Kitl are expressed in embryonic skin at appropriate times and sites to regulate melanoblast development ,,. In fact, sKitl enhances melanoblast proliferation to a greater extent when primary NC cells are cultured on chondroitin sulfate, suggesting that the ECM can modulate Kit signaling . Taken in context of this study, it could be melanoblasts in bt9J/bt9J explant cultures did not respond to sKitl because the ECM was defective.
Versican proteolysis could influence melanoblast behavior and Kit signaling by promoting direct interactions of Kitl with Kit receptor. Such a scenario is analogous to heparin sulfate proteoglycans, which modulate growth factor signaling by sequestering ligand and promoting its receptor binding . Studies of Weill-Marchesani syndrome strongly suggest ADAMTS10 (GeneID: 81794) regulates TGF βsignaling through its interactions with the ECM component, fibrillin-1 (GeneID: 2200) . Interestingly, versican binds extracellular cytokines, including the secreted growth factor, midkine (GeneID: 17242) –. Future studies will reveal if versican, Kit, and Kitl physically interact and if versican regulates Kit signaling directly.
Versican and its proteolytic products could also promote melanoblast survival through a pathway that is independent of Kit signaling. Versican V1 promotes survival of NIH3T3 cells and down-regulates expression of the pro-apoptotic protein Bad (GeneID: 12015) . Expression of the versican G1 domain protects sarcoma cells from apoptosis, and binding of the versican G3 domain to β-Integrin promotes survival of astrocytoma cells ,. Since integrins are expressed by melanoblasts ,, versican-integrin interactions are another potential mechanism by which versican could influence melanoblast development. Since versican mutants are embryonic lethal around E10.5 , conditional models of versican as well as transgenic animals containing constitutively cleaved and uncleaved Versican molecules will be necessary to assess the requirement of versican for melanoblast development.
ADAMTS20 and Human Disease
Given that cell survival is an integral component of melanoma progression, ADAMTS9 and ADAMTS20 are excellent candidates for participating in melanoma. ADAMTS20 over-expression has not been carefully examined in melanoma, but has been observed in brain, colon and breast tumors ,. The related metalloproteases ADAM9 (GeneID: 8754) and ADAMTS13 are upregulated in primary melanoma tumors and in melanoma cell lines, respectively ,. Interestingly, VERSICAN and KITL overexpression are observed in primary melanomas and levels correlate with melanoma progression –. Given the requirement of Adamts20 for Kit signaling and versican cleavage, it is intriguing to consider how dysregulation of signaling between ADAMTS20, Kit, and versican might contribute to melanoma progression.
Materials and Methods
Mice and Genetic Interactions
The following mouse stocks were used and kindly provided by: Adamts20bt-Bei1 (David Beier, Harvard Medical School), Adamts20bt9J (Lynn Lamoreux, Texas A&M University), Dct-LacZ, MiMi, Kittm1Alf, KitlSl (Heinz Arnheiter, NINDS), KitSl-d (Jackson Laboratories, Bar Harbor, ME). Adamts9ko and Adamts5ko knockout animals (both on a C57Bl/6 background) are described elsewhere (D. McCullough, H. Enomoto, S. Apte, unpublished). Quantification of white spotting was performed using Image J software (http://rsb.info.nih.gov/ij) (NIH, Bethesda, MD) and statistical significance of spotting calculated using ANOVA tests.
Genotyping and Sequencing
cDNA was sequenced by GeneDx (Gaithersburg, MD). Adamts20bt9J animals were genotyped using standard conditions on an ABIPrism7000 using a TaqmanTM assay of genomic DNA. For this assay, a region containing the point mutation is amplified by PCR. Two fluorescently labeled single-stranded oligonucleotides (probes), one complementary to the wild-type product, and one complementary to the mutant product are included in the assay. The relative amount of wild-type and mutant product is measured by fluorescence quenched upon DNA synthesis. Allelic discrimination was performed to detect the total levels of each allele at the conclusion of the PCR reaction. The following cycling conditions were used: 1. 95°C for 10 minutes 1×, 2. 92°C for 15 seconds, 3. 55°C for 60 seconds (2 and 3 repeated 40 times). The following primers and probes were used: TTCAGCACAGCTATTCTGGAAGAC (forward primer), GCACCTGAGGCAGACATACAC (reverse primer), CTTACCGAGATAGTTGTC (VIC probe, C57BL/6 allele), and CTTACCGAAATAGTTGTC (FAM probe, bt9J allele) (designed using Applied Biosystems software, Foster City, CA).
For quantitation of whole mount embryos, melanoblasts were counted from both sides of each embryo within a 1 mm×0.7 mm field surrounding the eye or the trunk. For quantitation of melanoblasts in E13.5 sections, Dct-LacZ expression was scored. The mitotic index was calculated using the average number of PH3+ melanoblasts per section divided by the average number of total melanoblasts per section and multiplied by 100. The apoptotic index was calculated using the average number of CC3+ melanoblasts per section divided by the average number of total melanoblasts per section and multiplied by 100. For quantitation of melanoblasts in different skin compartments the following criteria were used: melanoblasts above the basal membrane were classified as epidermal, melanoblasts below the basal membrane were classified as dermal, and melanoblasts crossing the basal membrane and present in both regions were classified as dermal/epidermal region. The following statistical tests were used: for average number of melanoblasts in sections and whole mount embryos (two-tailed Student's t-test), for proportion of total melanoblasts and apoptotic melanoblasts in skin layers (Chi-square test of independence), for proportion of proliferating and apoptotic melanoblasts (Fisher's test).
Whole Mount β-Galactosidase Staining and In situ Hybridization
β-galactosidase activity staining was performed as previously described . In situ hybridization was performed as previously described  using a 1 Kb mouse Adamts20 probe , a 644 bp mouse Versican V1 probe made using mouse Versican V1 cDNA (a kind gift of Andrew Copp, UCL), a Pmel17 probe made as previously described , and a 1 Kb mouse Adamts9 probe .
Frozen sections were prepared from embryos fixed overnight in 4% paraformaldehyde. Staining was performed with modifications to manufacturer's protocol (Vector Laboratories, Burlingame, CA), and sections were mounted using Hardset Vectashield with DAPI (Vector Laboratories). The following antibodies were used at 1∶200: mouse anti- β galactosidase (Promega, Madison, WI), rabbit anti-cleaved Caspase-3 (Cell Signaling, Danvers, MA), rabbit anti-phospho-Histone H3 (Upstate Biotechnology, Charlottesville, VA), rabbit anti-versican GAG β (Chemicon, Temecula, CA), rabbit anti-versican V0/V1 Neo (Affinity Bioreagents, Golden, CO), anti-mouse and anti-rabbit rhodamine-conjugated or FITC-conjugated secondary antibodies (Vector Laboratories).
Generation of Expression Plasmids and Versican Digestion
Adamts20 short isoform cDNA was assembled by PCR from E17.5 mouse embryo mRNA, sequence verified, and cloned into pcDNAmyc-hisA+ plasmid (Invitrogen, Thousand Oaks, CA). The expression plasmid for full-length Adamts9 was previously described . Transfected or untransfected cells were incubated with versican as previously described . Extracts of embryo skin were prepared using a RIPA lysis buffer containing protease inhibitor mixture (20 µl) and 2 mM phenylmethylsulphonyl fluoride (Pierce, Rockford, IL). For analysis of total versican, cell extracts were treated with 0.1 units chondroitinase ABC (Associates of Cape Cod, East Falmouth, MA) in 0.1 M Tris, 50 mM NaAc for 10 minutes at 37°C. Skin extracts and conditioned medium from transfected cells were run on 4–12% SDS-Polyacrylamide gels (Invitrogen) and analyzed by Western blotting using the following primary antibodies: mouse anti-myc 9E10 (Invitrogen), rabbit anti-versican GAG βat 1∶1000, rabbit anti-versican V0/V1 Neo at 1∶1000, and mouse anti-α-Tubulin at 1∶1000 (Sigma, St.Louis, MO); and secondary antibodies: anti-rabbit HRP and anti-mouse HRP (Amersham, Piscataway, NJ).
Ex vivo Embryonic Skin Analysis
Dorsal skin between the forelimb and hindlimb was isolated from E13.5 bt/+;Dct-LacZ and bt/bt;Dct-LacZ embryos with ventral tissue portions and internal organs removed. The skin explants were placed epidermal side up onto polyethylene terepthlate track-etched membranes in a cell culture insert (8.0 µm pore size, Becton-Dickinson, Franklin Lakes, NJ). Separate cultures were established for each explant. Explants were cultured for four days with 95% DMEM, 5% FBS (control medium) and as noted the medium was supplemented daily with 500 ng/ml Kitl (R&D Systems, Minneapolis, MN). After a culture period of four days, the explants were fixed in 4% paraformaldehyde in PBS (pH 7.4) for one hour and staining for β-galactosidase activity was performed as previously described .
Newborn Skin Analysis
Skin was removed from newborn pups, fixed in 4% paraformaldehyde overnight and washed 3× in PBS and then analyzed on the dermal side for pigmentation. Pigmentation was scored as the extent of pigmentation between anterior and posterior edges of samples, using the following scale: 1 (no pigmentation), 2 (between 0 and 25%), 3 (between 25 and 50%), 4 (between 50 and 75%), 5 (between 75% and 100%). Statistical significance was calculated using a student's t test.
Melanoblast distribution is normal in bt9J/bt9J E11.5 embryos. Images of β-galactosidase stained E11.5 bt9J/+;Dct-LacZ (A,C) and bt9J/bt9J;Dct-LacZ (B,D) embryos. Shown are whole embryos (A,B) and trunks (C,D). The scale bars are: (A,B) 2 mm, and (C,D) 1 mm.
(8.94 MB TIF)
Melanoblasts do not build up at the lateral edges of the belt in bt9J/bt9J E16.5 embryos. (A–D) Images of the belt region of β-galactosidase stained E16.5 whole mount bt9J/+;Dct-LacZ (A,C) and bt9J/bt9J;Dct-LacZ (B,D) embryos. Shown are low magnification (A,B) and high magnification(C,D) images from four different embryos. The scale bars are (A,C) 1 mm, and (B,D) 0.5 mm.
(10.38 MB TIF)
Kit heterozygous mutations do not exacerbate the bt phenotype at E12.5. Representative images of the trunks of E12.5 bt9J/+;Kittm1Alf/+ (A) and bt9J/bt9J;Kittm1Alf/+ embryos (B) (n = 6 each genotype). Melanoblasts are marked using LacZ, which is targeted to the Kittm1Alf locus. The scale bar is 1 mm.
(9.47 MB TIF)
The authors thank Tom Hornyak for Dct-LacZ DNA, Art Incao for technical assistance, Julia Fekes for graphical assistance, FEB of NCI and Laura Baxter for editorial assistance, Pam Schwartzberg and members of the Pavan laboratory for useful discussions and critical reading of the manuscript. Animal procedures were performed in accordance with protocol G94-7 from the NHGRI, NIH Institutional Review Board.
Conceived and designed the experiments: DS SA WP. Performed the experiments: DS LH RS MY. Analyzed the data: DS LH RS MY SA WP. Contributed reagents/materials/analysis tools: DS LH RS MY WP. Wrote the paper: DS SA WP.
- 1. Apte SS (2004) A disintegrin-like and metalloprotease (reprolysin type) with thrombospondin type 1 motifs: the ADAMTS family. Int J Biochem Cell Biol 36: 981–985.SS Apte2004A disintegrin-like and metalloprotease (reprolysin type) with thrombospondin type 1 motifs: the ADAMTS family.Int J Biochem Cell Biol36981985
- 2. Porter S, Clark IM, Kevorkian L, Edwards DR (2005) The ADAMTS metalloproteinases. Biochem J 386: 15–27.S. PorterIM ClarkL. KevorkianDR Edwards2005The ADAMTS metalloproteinases.Biochem J3861527
- 3. Glasson SS, Askew R, Sheppard B, Carito B, Blanchet T, et al. (2005) Deletion of active ADAMTS5 prevents cartilage degradation in a murine model of osteoarthritis. Nature 434: 644–648.SS GlassonR. AskewB. SheppardB. CaritoT. Blanchet2005Deletion of active ADAMTS5 prevents cartilage degradation in a murine model of osteoarthritis.Nature434644648
- 4. Le Goff C, Somerville RP, Kesteloot F, Powell K, Birk DE, et al. (2006) Regulation of procollagen amino-propeptide processing during mouse embryogenesis by specialization of homologous ADAMTS proteases: insights on collagen biosynthesis and dermatosparaxis. Development 133: 1587–1596.C. Le GoffRP SomervilleF. KestelootK. PowellDE Birk2006Regulation of procollagen amino-propeptide processing during mouse embryogenesis by specialization of homologous ADAMTS proteases: insights on collagen biosynthesis and dermatosparaxis.Development13315871596
- 5. Levy GG, Nichols WC, Lian EC, Foroud T, McClintick JN, et al. (2001) Mutations in a member of the ADAMTS gene family cause thrombotic thrombocytopenic purpura. Nature 413: 488–494.GG LevyWC NicholsEC LianT. ForoudJN McClintick2001Mutations in a member of the ADAMTS gene family cause thrombotic thrombocytopenic purpura.Nature413488494
- 6. Li SW, Arita M, Fertala A, Bao Y, Kopen GC, et al. (2001) Transgenic mice with inactive alleles for procollagen N-proteinase (ADAMTS-2) develop fragile skin and male sterility. Biochem J 355: 271–278.SW LiM. AritaA. FertalaY. BaoGC Kopen2001Transgenic mice with inactive alleles for procollagen N-proteinase (ADAMTS-2) develop fragile skin and male sterility.Biochem J355271278
- 7. Stanton H, Rogerson FM, East CJ, Golub SB, Lawlor KE, et al. (2005) ADAMTS5 is the major aggrecanase in mouse cartilage in vivo and in vitro. Nature 434: 648–652.H. StantonFM RogersonCJ EastSB GolubKE Lawlor2005ADAMTS5 is the major aggrecanase in mouse cartilage in vivo and in vitro.Nature434648652
- 8. Colige A, Sieron AL, Li SW, Schwarze U, Petty E, et al. (1999) Human Ehlers-Danlos syndrome type VII C and bovine dermatosparaxis are caused by mutations in the procollagen I N-proteinase gene. Am J Hum Genet 65: 308–317.A. ColigeAL SieronSW LiU. SchwarzeE. Petty1999Human Ehlers-Danlos syndrome type VII C and bovine dermatosparaxis are caused by mutations in the procollagen I N-proteinase gene.Am J Hum Genet65308317
- 9. Dagoneau N, Benoist-Lasselin C, Huber C, Faivre L, Megarbane A, et al. (2004) ADAMTS10 mutations in autosomal recessive Weill-Marchesani syndrome. Am J Hum Genet 75: 801–806.N. DagoneauC. Benoist-LasselinC. HuberL. FaivreA. Megarbane2004ADAMTS10 mutations in autosomal recessive Weill-Marchesani syndrome.Am J Hum Genet75801806
- 10. Tortorella MD, Burn TC, Pratta MA, Abbaszade I, Hollis JM, et al. (1999) Purification and cloning of aggrecanase-1: a member of the ADAMTS family of proteins. Science 284: 1664–1666.MD TortorellaTC BurnMA PrattaI. AbbaszadeJM Hollis1999Purification and cloning of aggrecanase-1: a member of the ADAMTS family of proteins.Science28416641666
- 11. Mayer TC, Maltby EL (1964) An Experimental Investigation of Pattern Development in Lethal Spotting and Belted Mouse Embryos. Dev Biol 22: 269–286.TC MayerEL Maltby1964An Experimental Investigation of Pattern Development in Lethal Spotting and Belted Mouse Embryos.Dev Biol22269286
- 12. Rao C, Foernzler D, Loftus SK, Liu S, McPherson JD, et al. (2003) A defect in a novel ADAMTS family member is the cause of the belted white-spotting mutation. Development 130: 4665–4672.C. RaoD. FoernzlerSK LoftusS. LiuJD McPherson2003A defect in a novel ADAMTS family member is the cause of the belted white-spotting mutation.Development13046654672
- 13. Schaible R (1972) Comparative effects of piebald-spotting genes on clones of melanocytes in different vertebrate species;. Riley V, editor. New York: Appleton-Century-Crofts. R. Schaible1972V. RileyComparative effects of piebald-spotting genes on clones of melanocytes in different vertebrate species;New YorkAppleton-Century-Crofts14
- 14. Lin JY, Fisher DE (2007) Melanocyte biology and skin pigmentation. Nature 445: 843–850.JY LinDE Fisher2007Melanocyte biology and skin pigmentation.Nature445843850
- 15. Silver D, Hou L, Pavan WJ (2006) The genetic regulation of pigment cell development;. Saint-Jeannet , editor. Georgetown: Landes Biosciences. pp. 155–169.D. SilverL. HouWJ Pavan2006Saint-JeannetThe genetic regulation of pigment cell development;GeorgetownLandes Biosciences155169
- 16. Serbedzija GN, Fraser SE, Bronner-Fraser M (1990) Pathways of trunk neural crest cell migration in the mouse embryo as revealed by vital dye labelling. Development 108: 605–612.GN SerbedzijaSE FraserM. Bronner-Fraser1990Pathways of trunk neural crest cell migration in the mouse embryo as revealed by vital dye labelling.Development108605612
- 17. Cable J, Jackson IJ, Steel KP (1995) Mutations at the W locus affect survival of neural crest-derived melanocytes in the mouse. Mech Dev 50: 139–150.J. CableIJ JacksonKP Steel1995Mutations at the W locus affect survival of neural crest-derived melanocytes in the mouse.Mech Dev50139150
- 18. Mackenzie MA, Jordan SA, Budd PS, Jackson IJ (1997) Activation of the receptor tyrosine kinase Kit is required for the proliferation of melanoblasts in the mouse embryo. Dev Biol 192: 99–107.MA MackenzieSA JordanPS BuddIJ Jackson1997Activation of the receptor tyrosine kinase Kit is required for the proliferation of melanoblasts in the mouse embryo.Dev Biol19299107
- 19. Rawls JF, Johnson SL (2003) Temporal and molecular separation of the kit receptor tyrosine kinase's roles in zebrafish melanocyte migration and survival. Dev Biol 262: 152–161.JF RawlsSL Johnson2003Temporal and molecular separation of the kit receptor tyrosine kinase's roles in zebrafish melanocyte migration and survival.Dev Biol262152161
- 20. Steel KP, Davidson DR, Jackson IJ (1992) TRP-2/DT, a new early melanoblast marker, shows that steel growth factor (c-kit ligand) is a survival factor. Development 115: 1111–1119.KP SteelDR DavidsonIJ Jackson1992TRP-2/DT, a new early melanoblast marker, shows that steel growth factor (c-kit ligand) is a survival factor.Development11511111119
- 21. Wehrle-Haller B, Weston JA (1995) Soluble and cell-bound forms of steel factor activity play distinct roles in melanocyte precursor dispersal and survival on the lateral neural crest migration pathway. Development 121: 731–742.B. Wehrle-HallerJA Weston1995Soluble and cell-bound forms of steel factor activity play distinct roles in melanocyte precursor dispersal and survival on the lateral neural crest migration pathway.Development121731742
- 22. Wehrle-Haller B, Meller M, Weston JA (2001) Analysis of melanocyte precursors in Nf1 mutants reveals that MGF/KIT signaling promotes directed cell migration independent of its function in cell survival. Dev Biol 232: 471–483.B. Wehrle-HallerM. MellerJA Weston2001Analysis of melanocyte precursors in Nf1 mutants reveals that MGF/KIT signaling promotes directed cell migration independent of its function in cell survival.Dev Biol232471483
- 23. Yoshida H, Kunisada T, Kusakabe M, Nishikawa S, Nishikawa SI (1996) Distinct stages of melanocyte differentiation revealed by analysis of nonuniform pigmentation patterns. Development 122: 1207–1214.H. YoshidaT. KunisadaM. KusakabeS. NishikawaSI Nishikawa1996Distinct stages of melanocyte differentiation revealed by analysis of nonuniform pigmentation patterns.Development12212071214
- 24. Matsui Y, Zsebo KM, Hogan BL (1990) Embryonic expression of a haematopoietic growth factor encoded by the Sl locus and the ligand for c-kit. Nature 347: 667–669.Y. MatsuiKM ZseboBL Hogan1990Embryonic expression of a haematopoietic growth factor encoded by the Sl locus and the ligand for c-kit.Nature347667669
- 25. Wehrle-Haller B, Morrison-Graham K, Weston JA (1996) Ectopic c-kit expression affects the fate of melanocyte precursors in Patch mutant embryos. Dev Biol 177: 463–474.B. Wehrle-HallerK. Morrison-GrahamJA Weston1996Ectopic c-kit expression affects the fate of melanocyte precursors in Patch mutant embryos.Dev Biol177463474
- 26. Yoshida H, Hayashi S, Shultz LD, Yamamura K, Nishikawa S, et al. (1996) Neural and skin cell-specific expression pattern conferred by steel factor regulatory sequence in transgenic mice. Dev Dyn 207: 222–232.H. YoshidaS. HayashiLD ShultzK. YamamuraS. Nishikawa1996Neural and skin cell-specific expression pattern conferred by steel factor regulatory sequence in transgenic mice.Dev Dyn207222232
- 27. Blelloch R, Anna-Arriola SS, Gao D, Li Y, Hodgkin J, et al. (1999) The gon-1 gene is required for gonadal morphogenesis in Caenorhabditis elegans. Dev Biol 216: 382–393.R. BlellochSS Anna-ArriolaD. GaoY. LiJ. Hodgkin1999The gon-1 gene is required for gonadal morphogenesis in Caenorhabditis elegans.Dev Biol216382393
- 28. Blelloch R, Kimble J (1999) Control of organ shape by a secreted metalloprotease in the nematode Caenorhabditis elegans. Nature 399: 586–590.R. BlellochJ. Kimble1999Control of organ shape by a secreted metalloprotease in the nematode Caenorhabditis elegans.Nature399586590
- 29. Hesselson D, Newman C, Kim KW, Kimble J (2004) GON-1 and fibulin have antagonistic roles in control of organ shape. Curr Biol 14: 2005–2010.D. HesselsonC. NewmanKW KimJ. Kimble2004GON-1 and fibulin have antagonistic roles in control of organ shape.Curr Biol1420052010
- 30. Hauschka TS, Jacobs BB, Holdridge BA (1968) Recessive yellow and its interaction with belted in the mouse. J Hered 59: 339–341.TS HauschkaBB JacobsBA Holdridge1968Recessive yellow and its interaction with belted in the mouse.J Hered59339341
- 31. Lamoreux ML (1999) Strain-specific white-spotting patterns in laboratory mice. Pigment Cell Res 12: 383–390.ML Lamoreux1999Strain-specific white-spotting patterns in laboratory mice.Pigment Cell Res12383390
- 32. Hornyak TJ, Hayes DJ, Chiu LY, Ziff EB (2001) Transcription factors in melanocyte development: distinct roles for Pax-3 and Mitf. Mech Dev 101: 47–59.TJ HornyakDJ HayesLY ChiuEB Ziff2001Transcription factors in melanocyte development: distinct roles for Pax-3 and Mitf.Mech Dev1014759
- 33. Wakamatsu Y, Mochii M, Vogel KS, Weston JA (1998) Avian neural crest-derived neurogenic precursors undergo apoptosis on the lateral migration pathway. Development 125: 4205–4213.Y. WakamatsuM. MochiiKS VogelJA Weston1998Avian neural crest-derived neurogenic precursors undergo apoptosis on the lateral migration pathway.Development12542054213
- 34. Baxter LL, Hou L, Loftus SK, Pavan WJ (2004) Spotlight on spotted mice: a review of white spotting mouse mutants and associated human pigmentation disorders. Pigment Cell Res 17: 215–224.LL BaxterL. HouSK LoftusWJ Pavan2004Spotlight on spotted mice: a review of white spotting mouse mutants and associated human pigmentation disorders.Pigment Cell Res17215224
- 35. Duttlinger R, Manova K, Chu TY, Gyssler C, Zelenetz AD, et al. (1993) W-sash affects positive and negative elements controlling c-kit expression: ectopic c-kit expression at sites of kit-ligand expression affects melanogenesis. Development 118: 705–717.R. DuttlingerK. ManovaTY ChuC. GysslerAD Zelenetz1993W-sash affects positive and negative elements controlling c-kit expression: ectopic c-kit expression at sites of kit-ligand expression affects melanogenesis.Development118705717
- 36. Pavan WJ, Tilghman SM (1994) Piebald lethal (sl) acts early to disrupt the development of neural crest-derived melanocytes. Proc Natl Acad Sci U S A 91: 7159–7163.WJ PavanSM Tilghman1994Piebald lethal (sl) acts early to disrupt the development of neural crest-derived melanocytes.Proc Natl Acad Sci U S A9171597163
- 37. Wilkie AL, Jordan SA, Jackson IJ (2002) Neural crest progenitors of the melanocyte lineage: coat colour patterns revisited. Development 129: 3349–3357.AL WilkieSA JordanIJ Jackson2002Neural crest progenitors of the melanocyte lineage: coat colour patterns revisited.Development12933493357
- 38. Ito M, Kawa Y, Ono H, Okura M, Baba T, et al. (1999) Removal of stem cell factor or addition of monoclonal anti-c-KIT antibody induces apoptosis in murine melanocyte precursors. J Invest Dermatol 112: 796–801.M. ItoY. KawaH. OnoM. OkuraT. Baba1999Removal of stem cell factor or addition of monoclonal anti-c-KIT antibody induces apoptosis in murine melanocyte precursors.J Invest Dermatol112796801
- 39. Bernex F, De Sepulveda P, Kress C, Elbaz C, Delouis C, et al. (1996) Spatial and temporal patterns of c-kit-expressing cells in WlacZ/+ and WlacZ/WlacZ mouse embryos. Development 122: 3023–3033.F. BernexP. De SepulvedaC. KressC. ElbazC. Delouis1996Spatial and temporal patterns of c-kit-expressing cells in WlacZ/+ and WlacZ/WlacZ mouse embryos.Development12230233033
- 40. Geissler EN, Ryan MA, Housman DE (1988) The dominant-white spotting (W) locus of the mouse encodes the c-kit proto-oncogene. Cell 55: 185–192.EN GeisslerMA RyanDE Housman1988The dominant-white spotting (W) locus of the mouse encodes the c-kit proto-oncogene.Cell55185192
- 41. Hou L, Panthier JJ, Arnheiter H (2000) Signaling and transcriptional regulation in the neural crest-derived melanocyte lineage: interactions between KIT and MITF. Development 127: 5379–5389.L. HouJJ PanthierH. Arnheiter2000Signaling and transcriptional regulation in the neural crest-derived melanocyte lineage: interactions between KIT and MITF.Development12753795389
- 42. Brannan CI, Lyman SD, Williams DE, Eisenman J, Anderson DM, et al. (1991) Steel-Dickie mutation encodes a c-kit ligand lacking transmembrane and cytoplasmic domains. Proc Natl Acad Sci U S A 88: 4671–4674.CI BrannanSD LymanDE WilliamsJ. EisenmanDM Anderson1991Steel-Dickie mutation encodes a c-kit ligand lacking transmembrane and cytoplasmic domains.Proc Natl Acad Sci U S A8846714674
- 43. Flanagan JG, Chan DC, Leder P (1991) Transmembrane form of the kit ligand growth factor is determined by alternative splicing and is missing in the Sld mutant. Cell 64: 1025–1035.JG FlanaganDC ChanP. Leder1991Transmembrane form of the kit ligand growth factor is determined by alternative splicing and is missing in the Sld mutant.Cell6410251035
- 44. Jordan SA, Jackson IJ (2000) MGF (KIT ligand) is a chemokinetic factor for melanoblast migration into hair follicles. Dev Biol 225: 424–436.SA JordanIJ Jackson2000MGF (KIT ligand) is a chemokinetic factor for melanoblast migration into hair follicles.Dev Biol225424436
- 45. Kashiwagi M, Kuroki T, Huh N (1997) Specific inhibition of hair follicle formation by epidermal growth factor in an organ culture of developing mouse skin. Dev Biol 189: 22–32.M. KashiwagiT. KurokiN. Huh1997Specific inhibition of hair follicle formation by epidermal growth factor in an organ culture of developing mouse skin.Dev Biol1892232
- 46. Kashiwagi M, Huh NH (2005) Organ culture of developing mouse skin and its application for molecular mechanistic studies of morphogenesis. Methods Mol Biol 289: 39–46.M. KashiwagiNH Huh2005Organ culture of developing mouse skin and its application for molecular mechanistic studies of morphogenesis.Methods Mol Biol2893946
- 47. Somerville RP, Longpre JM, Jungers KA, Engle JM, Ross M, et al. (2003) Characterization of ADAMTS-9 and ADAMTS-20 as a distinct ADAMTS subfamily related to Caenorhabditis elegans GON-1. J Biol Chem 278: 9503–9513.RP SomervilleJM LongpreKA JungersJM EngleM. Ross2003Characterization of ADAMTS-9 and ADAMTS-20 as a distinct ADAMTS subfamily related to Caenorhabditis elegans GON-1.J Biol Chem27895039513
- 48. Dutt S, Kleber M, Matasci M, Sommer L, Zimmermann DR (2006) Versican V0 and V1 guide migratory neural crest cells. J Biol Chem 281: 12123–12131.S. DuttM. KleberM. MatasciL. SommerDR Zimmermann2006Versican V0 and V1 guide migratory neural crest cells.J Biol Chem2811212312131
- 49. Henderson DJ, Ybot-Gonzalez P, Copp AJ (1997) Over-expression of the chondroitin sulphate proteoglycan versican is associated with defective neural crest migration in the Pax3 mutant mouse (splotch). Mech Dev 69: 39–51.DJ HendersonP. Ybot-GonzalezAJ Copp1997Over-expression of the chondroitin sulphate proteoglycan versican is associated with defective neural crest migration in the Pax3 mutant mouse (splotch).Mech Dev693951
- 50. Kubota Y, Morita T, Kusakabe M, Sakakura T, Ito K (1999) Spatial and temporal changes in chondroitin sulfate distribution in the sclerotome play an essential role in the formation of migration patterns of mouse neural crest cells. Dev Dyn 214: 55–65.Y. KubotaT. MoritaM. KusakabeT. SakakuraK. Ito1999Spatial and temporal changes in chondroitin sulfate distribution in the sclerotome play an essential role in the formation of migration patterns of mouse neural crest cells.Dev Dyn2145565
- 51. Perissinotto D, Iacopetti P, Bellina I, Doliana R, Colombatti A, et al. (2000) Avian neural crest cell migration is diversely regulated by the two major hyaluronan-binding proteoglycans PG-M/versican and aggrecan. Development 127: 2823–2842.D. PerissinottoP. IacopettiI. BellinaR. DolianaA. Colombatti2000Avian neural crest cell migration is diversely regulated by the two major hyaluronan-binding proteoglycans PG-M/versican and aggrecan.Development12728232842
- 52. Perris R, Perissinotto D, Pettway Z, Bronner-Fraser M, Morgelin M, et al. (1996) Inhibitory effects of PG-H/aggrecan and PG-M/versican on avian neural crest cell migration. Faseb J 10: 293–301.R. PerrisD. PerissinottoZ. PettwayM. Bronner-FraserM. Morgelin1996Inhibitory effects of PG-H/aggrecan and PG-M/versican on avian neural crest cell migration.Faseb J10293301
- 53. Carrino DA, Sorrell JM, Caplan AI (2000) Age-related changes in the proteoglycans of human skin. Arch Biochem Biophys 373: 91–101.DA CarrinoJM SorrellAI Caplan2000Age-related changes in the proteoglycans of human skin.Arch Biochem Biophys37391101
- 54. du Cros DL, LeBaron RG, Couchman JR (1995) Association of versican with dermal matrices and its potential role in hair follicle development and cycling. J Invest Dermatol 105: 426–431.DL du CrosRG LeBaronJR Couchman1995Association of versican with dermal matrices and its potential role in hair follicle development and cycling.J Invest Dermatol105426431
- 55. Sorrell JM, Carrino DA, Baber MA, Caplan AI (1999) Versican in human fetal skin development. Anat Embryol (Berl) 199: 45–56.JM SorrellDA CarrinoMA BaberAI Caplan1999Versican in human fetal skin development.Anat Embryol (Berl)1994556
- 56. Zimmermann DR, Dours-Zimmermann MT, Schubert M, Bruckner-Tuderman L (1994) Versican is expressed in the proliferating zone in the epidermis and in association with the elastic network of the dermis. J Cell Biol 124: 817–825.DR ZimmermannMT Dours-ZimmermannM. SchubertL. Bruckner-Tuderman1994Versican is expressed in the proliferating zone in the epidermis and in association with the elastic network of the dermis.J Cell Biol124817825
- 57. Sandy JD, Westling J, Kenagy RD, Iruela-Arispe ML, Verscharen C, et al. (2001) Versican V1 proteolysis in human aorta in vivo occurs at the Glu441-Ala442 bond, a site that is cleaved by recombinant ADAMTS-1 and ADAMTS-4. J Biol Chem 276: 13372–13378.JD SandyJ. WestlingRD KenagyML Iruela-ArispeC. Verscharen2001Versican V1 proteolysis in human aorta in vivo occurs at the Glu441-Ala442 bond, a site that is cleaved by recombinant ADAMTS-1 and ADAMTS-4.J Biol Chem2761337213378
- 58. Jungers KA, Le Goff C, Somerville RP, Apte SS (2005) Adamts9 is widely expressed during mouse embryo development. Gene Expr Patterns 5: 609–617.KA JungersC. Le GoffRP SomervilleSS Apte2005Adamts9 is widely expressed during mouse embryo development.Gene Expr Patterns5609617
- 59. Reymond A, Marigo V, Yaylaoglu MB, Leoni A, Ucla C, et al. (2002) Human chromosome 21 gene expression atlas in the mouse. Nature 420: 582–586.A. ReymondV. MarigoMB YaylaogluA. LeoniC. Ucla2002Human chromosome 21 gene expression atlas in the mouse.Nature420582586
- 60. Jordan SA, Jackson IJ (2000) A late wave of melanoblast differentiation and rostrocaudal migration revealed in patch and rump-white embryos. Mech Dev 92: 135–143.SA JordanIJ Jackson2000A late wave of melanoblast differentiation and rostrocaudal migration revealed in patch and rump-white embryos.Mech Dev92135143
- 61. Keshet E, Lyman SD, Williams DE, Anderson DM, Jenkins NA, et al. (1991) Embryonic RNA expression patterns of the c-kit receptor and its cognate ligand suggest multiple functional roles in mouse development. Embo J 10: 2425–2435.E. KeshetSD LymanDE WilliamsDM AndersonNA Jenkins1991Embryonic RNA expression patterns of the c-kit receptor and its cognate ligand suggest multiple functional roles in mouse development.Embo J1024252435
- 62. Yoshida H, Nishikawa S, Okamura H, Sakakura T, Kusakabe M (1993) The Role of c-Kit Proto-oncogene during Melanocyte Development in Mouse. In vivo Approach by the In utero Microinection of Anti-c-kit Antibody. Develop Growth & Differ 35: 209–220.H. YoshidaS. NishikawaH. OkamuraT. SakakuraM. Kusakabe1993The Role of c-Kit Proto-oncogene during Melanocyte Development in Mouse. In vivo Approach by the In utero Microinection of Anti-c-kit Antibody.Develop Growth & Differ35209220
- 63. Duttlinger R, Manova K, Berrozpe G, Chu TY, DeLeon V, et al. (1995) The Wsh and Ph mutations affect the c-kit expression profile: c-kit misexpression in embryogenesis impairs melanogenesis in Wsh and Ph mutant mice. Proc Natl Acad Sci U S A 92: 3754–3758.R. DuttlingerK. ManovaG. BerrozpeTY ChuV. DeLeon1995The Wsh and Ph mutations affect the c-kit expression profile: c-kit misexpression in embryogenesis impairs melanogenesis in Wsh and Ph mutant mice.Proc Natl Acad Sci U S A9237543758
- 64. Kluppel M, Nagle DL, Bucan M, Bernstein A (1997) Long-range genomic rearrangements upstream of Kit dysregulate the developmental pattern of Kit expression in W57 and Wbanded mice and interfere with distinct steps in melanocyte development. Development 124: 65–77.M. KluppelDL NagleM. BucanA. Bernstein1997Long-range genomic rearrangements upstream of Kit dysregulate the developmental pattern of Kit expression in W57 and Wbanded mice and interfere with distinct steps in melanocyte development.Development1246577
- 65. Kimura Y, Jones N, Kluppel M, Hirashima M, Tachibana K, et al. (2004) Targeted mutations of the juxtamembrane tyrosines in the Kit receptor tyrosine kinase selectively affect multiple cell lineages. Proc Natl Acad Sci U S A 101: 6015–6020.Y. KimuraN. JonesM. KluppelM. HirashimaK. Tachibana2004Targeted mutations of the juxtamembrane tyrosines in the Kit receptor tyrosine kinase selectively affect multiple cell lineages.Proc Natl Acad Sci U S A10160156020
- 66. Chesneau V, Becherer JD, Zheng Y, Erdjument-Bromage H, Tempst P, et al. (2003) Catalytic properties of ADAM19. J Biol Chem 278: 22331–22340.V. ChesneauJD BechererY. ZhengH. Erdjument-BromageP. Tempst2003Catalytic properties of ADAM19.J Biol Chem2782233122340
- 67. Cruz AC, Frank BT, Edwards ST, Dazin PF, Peschon JJ, et al. (2004) Tumor necrosis factor-alpha-converting enzyme controls surface expression of c-Kit and survival of embryonic stem cell-derived mast cells. J Biol Chem 279: 5612–5620.AC CruzBT FrankST EdwardsPF DazinJJ Peschon2004Tumor necrosis factor-alpha-converting enzyme controls surface expression of c-Kit and survival of embryonic stem cell-derived mast cells.J Biol Chem27956125620
- 68. Kawaguchi N, Horiuchi K, Becherer JD, Toyama Y, Besmer P, et al. (2007) Different ADAMs have distinct influences on Kit ligand processing: phorbol-ester-stimulated ectodomain shedding of Kitl1 by ADAM17 is reduced by ADAM19. J Cell Sci 120: 943–952.N. KawaguchiK. HoriuchiJD BechererY. ToyamaP. Besmer2007Different ADAMs have distinct influences on Kit ligand processing: phorbol-ester-stimulated ectodomain shedding of Kitl1 by ADAM17 is reduced by ADAM19.J Cell Sci120943952
- 69. Zou J, Zhu F, Liu J, Wang W, Zhang R, et al. (2004) Catalytic activity of human ADAM33. J Biol Chem 279: 9818–9830.J. ZouF. ZhuJ. LiuW. WangR. Zhang2004Catalytic activity of human ADAM33.J Biol Chem27998189830
- 70. Tajima Y, Moore MA, Soares V, Ono M, Kissel H, et al. (1998) Consequences of exclusive expression in vivo of Kit-ligand lacking the major proteolytic cleavage site. Proc Natl Acad Sci U S A 95: 11903–11908.Y. TajimaMA MooreV. SoaresM. OnoH. Kissel1998Consequences of exclusive expression in vivo of Kit-ligand lacking the major proteolytic cleavage site.Proc Natl Acad Sci U S A951190311908
- 71. Kunisada T, Yoshida H, Yamazaki H, Miyamoto A, Hemmi H, et al. (1998) Transgene expression of steel factor in the basal layer of epidermis promotes survival, proliferation, differentiation and migration of melanocyte precursors. Development 125: 2915–2923.T. KunisadaH. YoshidaH. YamazakiA. MiyamotoH. Hemmi1998Transgene expression of steel factor in the basal layer of epidermis promotes survival, proliferation, differentiation and migration of melanocyte precursors.Development12529152923
- 72. Wehrle-Haller B, Weston JA (1999) Altered cell-surface targeting of stem cell factor causes loss of melanocyte precursors in Steel17H mutant mice. Dev Biol 210: 71–86.B. Wehrle-HallerJA Weston1999Altered cell-surface targeting of stem cell factor causes loss of melanocyte precursors in Steel17H mutant mice.Dev Biol2107186
- 73. Takano N, Kawakami T, Kawa Y, Asano M, Watabe H, et al. (2002) Fibronectin combined with stem cell factor plays an important role in melanocyte proliferation, differentiation and migration in cultured mouse neural crest cells. Pigment Cell Res 15: 192–200.N. TakanoT. KawakamiY. KawaM. AsanoH. Watabe2002Fibronectin combined with stem cell factor plays an important role in melanocyte proliferation, differentiation and migration in cultured mouse neural crest cells.Pigment Cell Res15192200
- 74. Schlessinger J, Lax I, Lemmon M (1995) Regulation of growth factor activation by proteoglycans: what is the role of the low affinity receptors? Cell 83: 357–360.J. SchlessingerI. LaxM. Lemmon1995Regulation of growth factor activation by proteoglycans: what is the role of the low affinity receptors?Cell83357360
- 75. Hirose J, Kawashima H, Yoshie O, Tashiro K, Miyasaka M (2001) Versican interacts with chemokines and modulates cellular responses. J Biol Chem 276: 5228–5234.J. HiroseH. KawashimaO. YoshieK. TashiroM. Miyasaka2001Versican interacts with chemokines and modulates cellular responses.J Biol Chem27652285234
- 76. Wu YJ, La Pierre DP, Wu J, Yee AJ, Yang BB (2005) The interaction of versican with its binding partners. Cell Res 15: 483–494.YJ WuDP La PierreJ. WuAJ YeeBB Yang2005The interaction of versican with its binding partners.Cell Res15483494
- 77. Zou K, Muramatsu H, Ikematsu S, Sakuma S, Salama RH, et al. (2000) A heparin-binding growth factor, midkine, binds to a chondroitin sulfate proteoglycan, PG-M/versican. Eur J Biochem 267: 4046–4053.K. ZouH. MuramatsuS. IkematsuS. SakumaRH Salama2000A heparin-binding growth factor, midkine, binds to a chondroitin sulfate proteoglycan, PG-M/versican.Eur J Biochem26740464053
- 78. Sheng W, Wang G, Wang Y, Liang J, Wen J, et al. (2005) The roles of versican V1 and V2 isoforms in cell proliferation and apoptosis. Mol Biol Cell 16: 1330–1340.W. ShengG. WangY. WangJ. LiangJ. Wen2005The roles of versican V1 and V2 isoforms in cell proliferation and apoptosis.Mol Biol Cell1613301340
- 79. Wu Y, Chen L, Zheng PS, Yang BB (2002) beta 1-Integrin-mediated glioma cell adhesion and free radical-induced apoptosis are regulated by binding to a C-terminal domain of PG-M/versican. J Biol Chem 277: 12294–12301.Y. WuL. ChenPS ZhengBB Yang2002beta 1-Integrin-mediated glioma cell adhesion and free radical-induced apoptosis are regulated by binding to a C-terminal domain of PG-M/versican.J Biol Chem2771229412301
- 80. Wu Y, Chen L, Cao L, Sheng W, Yang BB (2004) Overexpression of the C-terminal PG-M/versican domain impairs growth of tumor cells by intervening in the interaction between epidermal growth factor receptor and beta1-integrin. J Cell Sci 117: 2227–2237.Y. WuL. ChenL. CaoW. ShengBB Yang2004Overexpression of the C-terminal PG-M/versican domain impairs growth of tumor cells by intervening in the interaction between epidermal growth factor receptor and beta1-integrin.J Cell Sci11722272237
- 81. Haass NK, Smalley KS, Li L, Herlyn M (2005) Adhesion, migration and communication in melanocytes and melanoma. Pigment Cell Res 18: 150–159.NK HaassKS SmalleyL. LiM. Herlyn2005Adhesion, migration and communication in melanocytes and melanoma.Pigment Cell Res18150159
- 82. Zambruno G, Marchisio PC, Melchiori A, Bondanza S, Cancedda R, et al. (1993) Expression of integrin receptors and their role in adhesion, spreading and migration of normal human melanocytes. J Cell Sci 105 (Pt 1): 179–190.G. ZambrunoPC MarchisioA. MelchioriS. BondanzaR. Cancedda1993Expression of integrin receptors and their role in adhesion, spreading and migration of normal human melanocytes.J Cell Sci105 (Pt 1)179190
- 83. Mjaatvedt CH, Yamamura H, Capehart AA, Turner D, Markwald RR (1998) The Cspg2 gene, disrupted in the hdf mutant, is required for right cardiac chamber and endocardial cushion formation. Dev Biol 202: 56–66.CH MjaatvedtH. YamamuraAA CapehartD. TurnerRR Markwald1998The Cspg2 gene, disrupted in the hdf mutant, is required for right cardiac chamber and endocardial cushion formation.Dev Biol2025666
- 84. Llamazares M, Cal S, Quesada V, Lopez-Otin C (2003) Identification and characterization of ADAMTS-20 defines a novel subfamily of metalloproteinases-disintegrins with multiple thrombospondin-1 repeats and a unique GON domain. J Biol Chem 278: 13382–13389.M. LlamazaresS. CalV. QuesadaC. Lopez-Otin2003Identification and characterization of ADAMTS-20 defines a novel subfamily of metalloproteinases-disintegrins with multiple thrombospondin-1 repeats and a unique GON domain.J Biol Chem2781338213389
- 85. Cal S, Obaya AJ, Llamazares M, Garabaya C, Quesada V, et al. (2002) Cloning, expression analysis, and structural characterization of seven novel human ADAMTSs, a family of metalloproteinases with disintegrin and thrombospondin-1 domains. Gene 283: 49–62.S. CalAJ ObayaM. LlamazaresC. GarabayaV. Quesada2002Cloning, expression analysis, and structural characterization of seven novel human ADAMTSs, a family of metalloproteinases with disintegrin and thrombospondin-1 domains.Gene2834962
- 86. Zigrino P, Mauch C, Fox JW, Nischt R (2005) Adam-9 expression and regulation in human skin melanoma and melanoma cell lines. Int J Cancer 116: 853–859.P. ZigrinoC. MauchJW FoxR. Nischt2005Adam-9 expression and regulation in human skin melanoma and melanoma cell lines.Int J Cancer116853859
- 87. Domenzain C, Docampo MJ, Serra M, Miquel L, Bassols A (2003) Differential expression of versican isoforms is a component of the human melanoma cell differentiation process. Biochim Biophys Acta 1642: 107–114.C. DomenzainMJ DocampoM. SerraL. MiquelA. Bassols2003Differential expression of versican isoforms is a component of the human melanoma cell differentiation process.Biochim Biophys Acta1642107114
- 88. Giehl KA, Nagele U, Volkenandt M, Berking C (2007) Protein expression of melanocyte growth factors (bFGF, SCF) and their receptors (FGFR-1, c-kit) in nevi and melanoma. J Cutan Pathol 34: 7–14.KA GiehlU. NageleM. VolkenandtC. Berking2007Protein expression of melanocyte growth factors (bFGF, SCF) and their receptors (FGFR-1, c-kit) in nevi and melanoma.J Cutan Pathol34714
- 89. Serra M, Rabanal RM, Miquel L, Domenzain C, Bassols A (2004) Differential expression of CD44 in canine melanocytic tumours. J Comp Pathol 130: 171–180.M. SerraRM RabanalL. MiquelC. DomenzainA. Bassols2004Differential expression of CD44 in canine melanocytic tumours.J Comp Pathol130171180
- 90. Touab M, Arumi-Uria M, Barranco C, Bassols A (2003) Expression of the proteoglycans versican and mel-CSPG in dysplastic nevi. Am J Clin Pathol 119: 587–593.M. TouabM. Arumi-UriaC. BarrancoA. Bassols2003Expression of the proteoglycans versican and mel-CSPG in dysplastic nevi.Am J Clin Pathol119587593
- 91. Touab M, Villena J, Barranco C, Arumi-Uria M, Bassols A (2002) Versican is differentially expressed in human melanoma and may play a role in tumor development. Am J Pathol 160: 549–557.M. TouabJ. VillenaC. BarrancoM. Arumi-UriaA. Bassols2002Versican is differentially expressed in human melanoma and may play a role in tumor development.Am J Pathol160549557
- 92. Baxter LL, Pavan WJ (2003) Pmel17 expression is Mitf-dependent and reveals cranial melanoblast migration during murine development. Gene Expr Patterns 3: 703–707.LL BaxterWJ Pavan2003Pmel17 expression is Mitf-dependent and reveals cranial melanoblast migration during murine development.Gene Expr Patterns3703707